PDCA and the LAIR Model

From SystemsWiki

Jump to: navigation, search

The PDCA (plan-do-check-act) is an iterative problem-solving process made popular by Dr. W. Edwards Deming. The cycle is typically used in business/organizational development processes and is referred to as the Deming cycle, Shewhart cycle, Deming wheel, or plan-do-study-act. When viewed from a systemic perspective it becomes evident that an evolution of the PDCA cycle would make it even more valuable in terms of fostering organizational development.

Contents

PDCA Shewhart Cycle

Fig. 1 - PDCA Shewhart Cycle

The PDCA Cycle is an iterative process where:

  • Plan. Establish objectives and process for attaining those objectives, integrating the fact that this is an iterative process.
  • Do. Implement the new process.
  • Check. Measure new process against expected results to ascertain variance.
  • Act. Determine cause of variance, determine changes for improvement, and repeat the sequence.

PDCA Systemic View

Fig. 2 is a presentation of the PDCA Shewhart Cycle as a Causal Loop Diagram.

Fig. 2 - PDCA as a Causal Loop Diagram [pdca-lair.msys]

The plan is intended to guide the migration of the current state from where it starts to the plan state. The difference between the two initial states is represented by the gap. This gap forms the basis for do which is the action to change the current state. Realizing that this is an iterative process, for each cycle the current state provides input to the check process which then results in an awareness of variance from what was expected and act is the alteration of the plan for the next iteration.

PDCA Evolution - The LAIR Model

Fig. 3 depicts a Causal Loop Diagram for an evolution of the PDCA Cycle into the LAIR model which stands for Learning through Action, Improvement and Rethinking.

Fig. 3 - LAIR Causal Loop Diagram [pdca-lair.msys]

The LAIR model is similar to the PDCA Cycle though makes the continued development more explicit and accommodates short, mid-term and long-term progress.

  • Action-Loop (B1) Represents the short term actions necessary to move the current state to the desired state and consists of actions based primarily on prior learning.
  • Improve-Loop (R2) Represents learning from the short-term actions and devising mid-term actions to actually extend the desired state beyond where it was, thus promoting another level of progress within the cycle.
  • Rethink-Loop (R3) Represents learning from short-term and mid-term actions and actually rethinking the value creation premise and thus extending the desired state even further for the long-term health of the business and possibly avoiding the myopic demise [Levitt 1975] which has been the reward for so many organizations that failed to rethink their business.

Another Perspective

For me, Nicolas Stampf, the check part is inherently a reflexive analysis of the experiment (the Do) where you should be learning. Learning thus tends to push you to improve the "desired state" and make you understand more of the system you're trying to improve (thus improving your ability to Plan and Do some new things). The Act part should be a part during which, depending on your learnings during Check, you decide whether or not to generalize the result of the experiment (in Lean it's the standardization).

Anyhow, Fig. 4 is the diagram I came to.

Fig. 4 - Stampf PDCA Model (PDCA_Stampf.mdl)

References

Additional Resources
Systems Thinking World Discussions
Systems Thinking World Q&A * Gene Bellinger
Retrieved from "http://systemswiki.org/index.php?title=PDCA_and_the_LAIR_Model"
Personal tools